a PDI=l.71. The PDI was significantly reduced due to higher initiation of this complex compared to 1. Treatment of 2 with 5.0 eq of 12 using the standard procedure resulted in complete initiation compared to the 15% initiation observed with 1.
Using complex 2, polymerization appeared possible to produce low polydisperse polymers without the need for the chelated monophosphines propagating species.
Therefore, a series of compounds was synthesized which did not contain the ether linkage in the side-chain (Scheme 11 ). The removal of the ether group should have reduced the
Cl
; ] ' O H
3
~COOMe
Cl
21
Na OH THF/H20 (1 :1)
98%
~COOH
24
71
amount of chelated monophosphine propagating species present during the polymerization, thus increasing the propagation rates and hopefully minimizing any decomposition side reactions. Oxidation of 3 with PCC46 resulted in moderate yield of cis,trans-3-
chlorocyclobutanecarboxaldehyde (20). This was subjected to a Homer-Emmons reaction47,48 with trimethyl 4-phosphonocrotonate using the method of Roush49 to generate 21. Hydrogenation of the diene with Pd/C and subsequent saponification of the ester produced nearly quantitative yield of 23. Elimination using the standard conditions resulted in 5-(2-cyclobutenyl)pentanoic acid (24) in 92% yield. Conversion of 24 into a series of functionalized cyclobutenes was then attempted (Scheme 12). Reduction of 24
Scheme 12
~CONHBu 1) SOCl2
[d 2) H2NBu, THF
26 85%
~COOMe 27
K2C03, Mel Acetone 99%
~COOH 24
LAH THF 99%
~OH
25
with LAH produced 5-(2-cyclobutenyl)pentanol (25) in quantitative yield. Treatment of 24 with thionyl chloride followed by reaction with n-butylamine resulted in N-butyl-[5-(2- cyclobutenyl)]pentanamide (26) in 85% yield. Finally, alkylation of the carboxylic acid 24 with K1C03 and Mel gave a quantitative yield of the methyl [5-(2-cyclobutenyl)]pentanoate (27). These four monomers 24-27 contained a variety of functional groups; however, with the functional group well removed from the ring system, it was predicted that the amount of chelation would be similar and the monomers would react at similar rates.
Optimal polymerization conditions were at room temperature for 1.5 h with a [M]=0.17 Musing initiator 2. A slight reduction in the PDis was observed at lower [M]
resulting in the optimum concentration at 0.17 M. THF was chosen as a solvent because the polymers of the acid and alcohol monomers 24 and 25 were insoluble in all other solvents compatible with 2. As seen in Table 4, complete initiation was observed
employing 5 eq of monomers 24-27 with 2. Analysis of the carbene propagating species by 1 H and 31 P NMR was then undertaken. The 1 H NMR for the polymerization of 24-27 were similar in the four cases, so only the results for 24 will be discussed in detail (Table 6). Two species at 19.49 and 17.90 ppm were observed in a ratio of 1.00:0.41. 31p NMR
Table 6. 1 H NMR analysis of polymerization propagating species
x
% lHNMR lHNMR Ratio (103Keq)bInitiation Carbene Carbene B/Ab with 5 Resonance Resonance
eq. [MP of Ab ofBb
(CH2hCOOH >99 % 19.49 17.90 0.41 1.2
(CH2)40H >99% 19.49 17.79 0.57 2.1
(CH2hC(O)NHBu >99 % 19.47 17.89 0.35 0.89
(CH2)3COOMe >99% 19.49 17.89 0.51 1.7
a Polymerizations were run in THF-dg at room temperature for 1 h at [M]=0.11 M and [1]=0.022 M and analyzed by 1 H NMR using a JEOL GX-400 spectrometer. b Polymerizations were run in THF-dg at room temperature for 1 h at [M]=0.17 M and [1]=0.0099 M and analyzed by 1 H NMR using a JEOL GX-400 spectrometer.
73
displayed nearly identical spectra for the four cases and the results for 24 were
representative with a multiplet at 38.58 ppm for the chelated species and a series of peaks from 37.11-34.19 ppm for the bisphosphine species. Free PCy3 was found at 10.80 ppm integrating 1.00: 1.02 with the multiplet at 38.58. So, as expected, the bisphosphine species was favored over the chelated monophosphine species and the observed ratios were similar for all four compounds (Table 6). In contrast to the polymerization of 17, free phosphine was present in the polymerization of 24, and no decomposition products were observed. To further examine this, PCy3 was reacted for 1 h with 24 under the conditions used in the polymerization at room temperature in THF with a [24]=0.17 M and a
[PCy3]=0.0l M. Unlike for 17, no reaction was observed. To better compare the two cases, the reaction was repeated using the same conditions as for 17 at 45°C with a [24]=0.57 Mand a [PCy3]=0.01 M. In this case, only a minor species was observed at 52.82 integrating to 8.9% compared with free PCy3. The marked difference in reactivity of 17 and 24 with free PCy3 was proposed to be the result simply of the difference in acidity expected for these carboxylic acids. Using pentanoic acid and methoxyacetic acid as
models for 24 and 17, the pKa values are 4.80 and 3.5750 providing a plausible reason for reduced protonation of PCy3 by 24. The lack of reactivity of 24 with PCy3 was one explanation for the successful polymerization of 24 compared to the decomposition found in the polymerization of 17. Further study of these polymerization is necessary to
conclusively determine the cause of these contrasting polymerization results.
In order to fully investigate these polymerizations, the dependence of the Mn on the
[M]/[I] ratio was studied. The dependence was linear for all four monomers providing for control over the polymer molecular w~ight, and the polymers produced were of low polydispersity between 1.11-1.20 in all cases (Table 7). Figure 3 illustrates the dependence for the polymerization of the carboxylic acid 24. The linear dependence observed was supportive of a living polymerization, 14• 15 but further proof was necessary.
74
In addition to the molecular weight study, evidence for the lack of chain transfer and chain termination reactions must be demonstrated to prove that a polymerization is living. A
Table 7. Polymerization results for monomers 24-27a
Entry Monomer [M]/[I] - b
Mn PDib
1 24 25.0 5000 1.15
2 24 50.0 9200 1.16
3 24 75.0 12400 1.16
4 24 104 15700 1.16
5 24 150 20300 1.19
6 25 25.0 4200 1.15
7 25 50.4 7400 1.15
8 25 75.6 10100 1.17
9 25 101 13800 1.15
10 25 161 23300 1.20
11 26 25.2 4800 1.11
12 26 50.4 8800 1.11
13 26 73.3 11400 1.11
14 26 105 16400 1.12
15 26 151 20800 1.12
16 27 25.3 4600 1.14
17 27 50.5 8000 1.14
18 27 73.5 11800 1.13
19 27 106 16100 1.15
20 27 143 23100 1.16
a Polymerizations were run in THF at rt with [M]=0.17 M. b Determined by gel permeation chromatography in THF relative to monodispersed polystyrene standards.
19560
16300
~= 13040
9780
6520
75
40 60 80 100 120 140 160 [M]/[C]
Figure 3. Molecular weight dependence of the polymerization of 24 on the [M]/[I].
sequential monomer addition experiment was proposed to address the living nature of these polymerizations.14 This experiment was run for all four monomers, but the results for the polymerization of carboxylic acid 24 were representative. Using initiator 2, 25 eq of 24 were polymerized for 2 h under the standard conditions. The solution was then divided into three portions. Portion A was removed and analyzed by GPC. Portion B was left stirring for an additional 4 h. To portion C, an additional 350 eq of monomer were added, and this was left to react for 4 h. Portions B and C were then analyzed by GPC. As seen in Figure 4, no significant change was observed from the 2 h reaction (A) and the 6 h reaction (B). After 2 h, the Mn=4400 and PDl=l.14, and after 6 h, the Mn=4400 and PDl=l.14. If any significant chain transfer processes had been operative during this time, broadening of the PDI would have been observed. Portion C in which 350 additional eq of 24 had been polymerized resulted in a Mn =53400 and a PDl=l.34. Despite the fact that the PDI broadened, a clean shift in the GPC peaks was accomplished, proving that no
chain termination processes were occurring. Chain termination would have resulted in inactive chain ends after the polymerization of the first 25 eq monomer which would have produced a bimodal GPC trace after additional polymerization. The absence of both chain transfer and chain termination reactions was indicative of a living polymerization of the monomers 24-27.
12 14
,.,
I \ I \
I \
16
\ I
18 20 GPC Elution, cm
- - - A o B ---C
22 24
Figure 4. Sequential monomer addition experiment with A) 25 eq of 24 after 2 h, B) 25 eq of 24 after 6 h, and C) condition A after 350 additional eq of 24.
Polymer characterization. Through polymerization of these substituted cyclobutenes, a new group of poly(butadiene )s bearing a wide range of functional groups were prepared (Table 8). Even though the polymerization yields were quantitative by 1 H NMR, isolated yields were lower due to loss in the purification step. No significant bias in the olefin configuration was evident for these polymerizations with between 40 to 50% cis- olefin present in the polymer backbone. These results were consistent with that seen
77
previously for initiators 1and2.28 An accurate determination was impossible for poly(lO) and poly(lS) due to the lack of resolution of the olefinic resonances in the 1 H NMR
Table 8. Polymerization results
Pol~mer
x
% Yielda % Cis-OlefinbPoly(8) OCH2Ph 87 40
Poly(lO) OC(Phh 91 ndC
Poly(l2) OC(O)Ph 95 50
Poly(15) N(i-Pr)i 79 ndC
Poly(l8) OCH2COOMe 84 40
Poly(24) (CH2)3COOH 93 50
Poly(25) (CH2)40H 95 50
Poly(26) (CH2)3CONHBu 92 50
Poly(27) (CH2)3COOMe 96 50
a Isolated yields after purification. b Determined by 1 H NMR integration of the olefinic resonances. c Not determined; the olefinic resonances were not resolved well enough to determine this value accurately.
spectrum. Based on analysis of the olefin region of the
Be
NMR spectrum of the polymers, a lack of regioselectivity for monomer insertion in the polymerizations was determined,51 again consistent with previous observations with these initiators in similar systems.28 If chelation occurs during the initial monomer coordination step, one might expect some regioselectivity due to energetic differences in the two possible chelate ring sizes. The lack of regioselectivity observed in these polymerizations is thus supportive of the proposal that chelation occurs after the ring-opening metathesis step and therefore plays no role in affecting the regioselectivity of monomer insertion.Analysis of the thermal properties of these polymers was undertaken as well using both differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) (Table 9). The Tg values for these polymers ranged from -46.6 to 69.4°C demonstrating the
Table 9. Polymer thermal behavior
Polymer
x
T g (l)(°C)a Tct (Argon) (°C)b T ct (Air) (°C)bPoly(8) OCH2Ph -22.6 306 254
Poly(lO) OC(Phh 69.4 308 240
Poly(12) OC(O)Ph 24.0 349 341
Poly(lS) N(i-Pr)i -15.8 294 249
Poly(18) OCH2COOMe -42.9 334 311
Poly(24) (CH2hCOOH 1.3 352 359
Poly(25) (CH2)40H -37.1 220 265
Poly(26) (CH2hCONHBu 0.1 283 259
Poly(27) (CH2hCOOMe -46.6 345 286
a Analysis by differential scanning calorimetry with a scan rate of l0°C/min. b Analysis by thermal gravimetric analysis with a scan rate of l0°C/min.
striking influence of the side-chains on the phase transitions of these poly(butadiene )s;
however, no melting transitions were observed for any of the materials. TGA results were compared using the T ct, the temperature at 10% decomposition. T ct values between 220 and 349°C were observed in an inert atmosphere, but no consistent trend was evident.
Diblock copolymerization experiments. The ability to prepare well-defined block copolymers has been realized through the development of living polymerization methods.15 Such polymeric materials have found application for many types of materials including compatiblizers for polymer blends, thermoplastic elastomers, surface-active agents, and semipermeable membranes.15,52,53 The attainment of living polymerizations
of monomers 24-27 using initiator 2 prompted an investigation of the synthesis of diblock copolymers using these monomers. Highly functionalized diblock copolymers with acid, ester, amide, and alcohol functionalities were envisioned through these di block
copolymerization experiments.
In these diblock copolymerization experiments, initiator 2 was treated with 50 eq of the first monomer for 1.5 h. At the end of this reaction time, 50 eq of the second monomer was then added, and the reaction was allowed to proceed at room temperature for an
additional 2 h (Scheme 13). The polymerization was then quenched with ethyl vinyl ether and purified as always to result in the diblock copolymer. Five different diblock
copolymers were prepared using this methodology in order to generate diblock copolymers with a wide range of functional groups. Ester-alcohol, ester-acid, amide-ester, amide-acid, and alcohol-acid diblock copolymers were synthesized. As seen in Table 10, the yields were excellent between 92-98%, and by lH NMR integration, a 1:1 ratio of the two polymer blocks was achieved in all cases. As previously observed in the
homopolymerizations of monomers 24-27, approximately 50% cis-olefin was obtained in all of the diblock copolymers. Although the polydispersities were slightly broader than in the homopolymerizations, low polydispersities between 1.17-1.24 were achieved
demonstrating the high degree of control in the living polymerizations of these functionalized cyclobutenes.